[Pellet-users] Examples of When Pellet Runs Slowly

[Rinke Hoekstra]

(apologies for the crosspost)

Hi Anthony, Kendall,

As I posted to the TB mailinglist some time ago. I experienced the  
same problem. This is usually caused by inconsistencies in the  
ontology you're trying to check (mind you, I haven't looked at your  
particular example).

The slowness is caused, as Holger also pointed out, by the particular  
way in which the Pellet-Jena bridge is implemented in TopBraid. This  
implementation deviates from the examples on the Pellet website  
because of particularities in the Eclipse plugin architecture (I don't  
know the details), and the retrieval of reasoner results (I guess).

If you run Pellet directly either through Jena (in your own code), or  
with the OWLAPI (e.g. Protege 4/SWOOP) performance should be much  
better (i.e. the usual max. of a couple of minutes vs. the hours you  
were talking about).

Good luck,

	Rinke


On 26 mrt 2008, at 13:33, JPQ wrote:
> > (I see that someone with the same email address posted about a  
> seemingly related issue yesterday, but that was about Pellet
> > semantics, not peformance, so I won't make any assumptions about a  
> connection.)
>
> That was me. In fact, at the very beginning of that email, I stated  
> the following. "This post was prompted by a post I made yesterday.  
> The setup is somewhat similar, but the subject matter is different."
>
> > We don't support TopBraid Composer on this list. You're better off  
> asking TopQuadrant about it.
>
> Granted. I had posted a couple of Pellet-related questions on the  
> TBC list in the past. They were addressed, but I was asked to post  
> them here. Thank you for answering the question. I'll certainly have  
> to look at Pellet and the OWLAPI independently.
>
> > Finally, a word about "best practices" in the context of this  
> mailing list: we are very much not willing to give free tech support  
> to people
> > who do not identify themselves by name in email.
>
> No disrespect intended.
>
>
> Cheers,
>
> Anthony
>
> On 3/25/08, Kendall Grant Clark <kendall at clarkparsia.com> wrote:  
> Excerpts from jpq.semantics's message of Tue Mar 25 15:39:40 -0400  
> 2008:
>
> > The items marked with asterisks indicate those variations which  
> had Pellet
> > running very slowly. Pellet (Pellet library plug-in, version  
> 1.5.1) was run
> > within TopBraid Composer (version 2.4.2). At most I waited  
> approximately one
> > hour to allow Pellet to complete its inferencing on the marked  
> items.
>
> We don't support TopBraid Composer on this list. You're better off
> asking TopQuadrant about it.
>
> The way they use Pellet internally -- via Jena, the last time we
> checked, rather than OWLAPI -- is not recommended or ideal, so we
> can't say anything about performance problems encountered when using
> Pellet in ways we don't recommend.
>
> If you can repeat yr observations with Pellet directly, we'd be happy
> to consider yr bug report, ideally with minimal code and ontology
> sufficient to demonstrate the problem. (I see that someone with the
> same email address posted about a seemingly related issue yesterday,
> but that was about Pellet semantics, not peformance, so I won't make
> any assumptions about a connection.)
>
> Finally, a word about "best practices" in the context of this mailing
> list: we are very much not willing to give free tech support to people
> who do not identify themselves by name in email. Other people on this
> list may be willing to help in that circumstance, but we aren't among
> them. FWIW. :>
>
> --
> Cheers,
> Kendall
>
> _______________________________________________
> Pellet-users mailing list
> Pellet-users at lists.owldl.com
> http://lists.owldl.com/mailman/listinfo/pellet-users
> _______________________________________________
>
> Sponsored by Clark & Parsia, LLC http://clarkparsia.com/

-----------------------------------------------
Drs. Rinke Hoekstra

Email: hoekstra at uva.nl    Skype:  rinkehoekstra
Phone: +31-20-5253499     Fax:   +31-20-5253495
Web:   http://www.leibnizcenter.org/users/rinke

Leibniz Center for Law,          Faculty of Law
University of Amsterdam,            PO Box 1030
1000 BA  Amsterdam,             The Netherlands
-----------------------------------------------